Advances, Systems and Applications
From: An overview of QoS-aware load balancing techniques in SDN-based IoT networks
No. Questions | Conclusion | ||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Question 1 (load balancing techniques) | QoS-aware load-balancing techniques based on maximum use of centralized and distributed architectures are shown in the table below. Also, the table number related to the characteristics of the technique is mentioned | ||||||||||||||||
Routing | Architecture | Offloading | Allocation | Classification | Clustering | Scheduling | Migration | Others | |||||||||
Centralized | ✓ | ✓ | ✓ | ✓ | ✓ | ||||||||||||
Distributed | ✓ | ✓ | ✓ | ✓ | |||||||||||||
No. Table | 2 | 4 | 3 | 7 | 5 | 9 | 8 | 6 | 10 | ||||||||
The following table shows the use of load-balancing techniques in SD-IoT network layers. In general, further need for load balancing at the fog/controller layer is discussed | |||||||||||||||||
Routing | Architecture | Offloading | Allocation | Classification | Clustering | Scheduling | Migration | Others | |||||||||
Application | ✓ | ✓ | |||||||||||||||
Control/fog | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ||||||||||
Infrastructure | ✓ | ✓ | ✓ | ||||||||||||||
Question 2 (QoS parameters) | The load balancing techniques that lead to the maximum improvement of QoS parameters are shown in the table below | ||||||||||||||||
QoS Techniques | Delay | Response time | Throughput | Resource Utilization | Cost | Packet loss ratio | Packet delivery ratio | Energy consumption | Security | Load balancing | |||||||
Routing | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | |||||||||||
Architecture | ✓ | ✓ | ✓ | ✓ | ✓ | ||||||||||||
Offloading | ✓ | ✓ | ✓ | ||||||||||||||
Allocation | ✓ | ✓ | ✓ | ||||||||||||||
Classification | ✓ | ||||||||||||||||
Clustering | ✓ | ||||||||||||||||
Scheduling | ✓ | ✓ | ✓ | ||||||||||||||
Migration | ✓ | ✓ | ✓ | ✓ | ✓ | ||||||||||||
Others | ✓ | ✓ | ✓ | ||||||||||||||
Table shows that most researchers have concentrated on delay, followed by response time, throughput, and resource utilization | |||||||||||||||||
Question 3 (multi-directions QoS) | As explained in Table 14, according to the QoS directions and SD-IoT network layers, three horizontal and vertical layers can be proposed, and nine parameters can be obtained from their combination | ||||||||||||||||
Directions layers | User | Service provider | Infrastructure provider | ||||||||||||||
Application/Cloud | Cost | Throughput | Resource utilization | ||||||||||||||
Control/Fog | Delay | Response time | Reliability | ||||||||||||||
Infrastructure | Accessibility | Energy consumption | Energy consumption | ||||||||||||||
Question 4 (applications) | Figure 11 shows that most researchers have concentrated on smart city applications, followed by large-scale, industry, and multimedia | ||||||||||||||||
Question 5 (simulation tools) | General and specialized simulation tools and real beds can be used for simulation. Most researchers have used Matlab and Mininet tools to evaluate their techniques. Figure 12 shows the evaluation tools used in different research articles | ||||||||||||||||
Question 6 (future research) | Future research is described in Sect. 6, which is divided into three general categories • Application-based load-balancing • Controller-Architecture-based load-balancing technique • Multi-constraint many-objective QoS in multi-direction |